How do I run the results commands from the shell?
Table of Contents
This file documents the results resource at the shell, one section per command. Every command is a subcommand of results, and every section exports its own script into projects/ores.shell/scripts/library/. Loading a script runs that one command, so a failure names the command it came from.
1. Commands
1.1. list
Reads one page of the collection. Nothing addresses it but the caller's tenant, so it is the read that shows what exists.
The command is shaped like this:
results list [--offset <n>] [--limit <n>] [--order <field>] [--desc]
It is addressed at compute.v1.results.list.
results list --limit 1
1.2. get
Reads the one row its key addresses. A key states every identifying column, so a partial key is refused rather than answered with an arbitrary row.
The command is shaped like this:
results get <id>
It is addressed at compute.v1.results.get.
results get 00000000-0000-0000-0000-000000000000
1.3. get-many
Reads several rows in one request, one key group per row.
The command is shaped like this:
results get-many <id>
It is addressed at compute.v1.results.get_many.
results get-many 00000000-0000-0000-0000-000000000000
1.4. add
Writes a row. The change states that the row must not exist, so a create that collides with an existing row is refused rather than replacing it. The change carries a reason and a commentary, which the server records on it so the row's history says why it moved.
The command is shaped like this:
results add <workunit_id> <host_id> <pgmq_msg_id> <server_state> <outcome> <output_uri> <error_message> <received_at> <reason> <commentary>
It is addressed at compute.v1.results.put.
results add 00000000-0000-0000-0000-000000000000 00000000-0000-0000-0000-000000000000 0 0 0 __none__ __none__ 1970-01-01T00:00:00Z system.new_record generated_script
1.5. set
Writes a row. The change states no precondition, so it replaces whatever row the key already addresses. The change carries a reason and a commentary, which the server records on it so the row's history says why it moved. Passing –version makes the write conditional on the version the caller last read.
The command is shaped like this:
results set <id> <workunit_id> <host_id> <pgmq_msg_id> <server_state> <outcome> <output_uri> <error_message> <received_at> <reason> <commentary> [--version <n>]
It is addressed at compute.v1.results.put.
results set 00000000-0000-0000-0000-000000000000 00000000-0000-0000-0000-000000000000 00000000-0000-0000-0000-000000000000 0 0 0 __none__ __none__ 1970-01-01T00:00:00Z system.new_record generated_script
1.6. put-many
Writes several rows in one request. The count states how many field groups follow. The change carries a reason and a commentary, which the server records on it so the row's history says why it moved.
The command is shaped like this:
results put-many --count <n> <id> <workunit_id> <host_id> <pgmq_msg_id> <server_state> <outcome> <output_uri> <error_message> <received_at> <reason> <commentary>
It is addressed at compute.v1.results.put_many.
results put-many --count 1 00000000-0000-0000-0000-000000000000 00000000-0000-0000-0000-000000000000 00000000-0000-0000-0000-000000000000 0 0 0 __none__ __none__ 1970-01-01T00:00:00Z system.new_record generated_script
1.7. delete
Removes a row. The change carries a reason and a commentary, which the server records on it so the row's history says why it moved. Passing –version makes the write conditional on the version the caller last read.
The command is shaped like this:
results delete <id> <reason> <commentary> [--version <n>]
It is addressed at compute.v1.results.delete.
results delete 00000000-0000-0000-0000-000000000000 system.new_record generated_script
1.8. delete-many
Removes several rows in one request. The change carries a reason and a commentary, which the server records on it so the row's history says why it moved.
The command is shaped like this:
results delete-many <id> <reason> <commentary>
It is addressed at compute.v1.results.delete_many.
results delete-many 00000000-0000-0000-0000-000000000000 system.new_record generated_script
1.9. by-workunit-id
Reads one page of the rows that share a relation value. The relation is part of the address rather than a filter applied after the read.
The command is shaped like this:
results by-workunit-id <workunit_id> [--offset <n>] [--limit <n>] [--order <field>] [--desc]
It is addressed at compute.v1.results.list_by_workunit_id.
results by-workunit-id 00000000-0000-0000-0000-000000000000 --limit 1
1.10. versions
Reads the row's recorded history, newest first.
The command is shaped like this:
results versions <id> [--offset <n>] [--limit <n>] [--order <field>] [--desc]
It is addressed at compute.v1.results_versions.list.
results versions 00000000-0000-0000-0000-000000000000 --limit 1
1.11. version
Reads one recorded version of the row.
The command is shaped like this:
results version <id> --version <n>
It is addressed at compute.v1.results_versions.get.
results version 00000000-0000-0000-0000-000000000000 --version 1